From the early days of silica optical fiber, it has been axiomatic to avoid using transmission wavelengths near 1383 nm the "terra incognita" of the fiber world. At this wavelength, the fiber's attenuation could be high and unreliable, and who knows what could happen if you unwittingly ventured there.
The culprit? Residual moisture in the fiber deposited during the manufacturing process. When incorporated into silica, the oxygen and hydrogen atoms in the hydroxyl "water" ion vibrate at 2.73 microns and produce a large absorption loss at this wavelength. Although 2.73 microns lies above the longest wavelength where fibers are used, one of the overtones of this wavelength produces the "water-peak" loss near 1383 nm. Depending on the magnitude of this peak, wavelengths within ±50 nm of the peak can be affected.
Some standards bodies have ignored the water-peak region altogether. For example, the International Electrotechnical Commission (IEC) and the International Telecommunication Union (ITU) place no restriction on a fiber's water-peak attenuation, whereas the Telecommunications Industry Association (TIA) allows it to be as high as 2.1 dB/km. With specifications this lax, it's not surprising that transmission-equipment manufacturers have avoided wavelengths near 1383 nm. This attitude is expected to change as standards for low-water-peak fibers remove the obstacle to a seamless optical spectrum extending from 1260 to 1625 nm.
The challenge in writing a standard for low-water-peak fiber is to ensure that such a fiber not only has a low initial water peak, but also that the peak remains stable with time. When used in the usual 1310- and 1550-nm operating wavelength windows, standards bodies have generally concluded that the amount of hydrogen present in cables is so small that its contribution to attenuation is insignificant. At 1383 nm, how ever, Kai Chang and coworkers at Lucent Technologies reported at OFC '99 that the effects of even small hydrogen levels can accumulate with time and that the added permanent attenuation can become significant.
At its June meeting, TIA FO-6.6, chaired by Greg Smith of Corning Inc., approved sending out for letter ballot a draft detail specification for low-water-peak dispersion-unshifted singlemode fiber. Prepared by yours truly, the draft contains an accelerated hydrogen aging test. It treats a low-water-peak fiber as one whose average attenuation at 1383 nm after hydrogen aging does not exceed the customer-specified attenuation at 1310 nm. This characteristic gives assurance that the fiber will maintain equivalent functionality at the two wavelengths over its installed service life.
In addition to activity in TIA, the United States included a low-water-peak fiber standard as part of its proposed restructuring of the 15-year-old ITU Recommendation G.652 for dispersion-unshifted singlemode fiber. Presented at the ITU-T Study Group 15 Question 15 meeting in June, this proposal will be discussed again at an October interim meeting. Next stop is a submission, also in October, to IEC Working Group 1 of Subcommittee 86A.
James J. Refi is with Lucent Technologies (Norcross, GA) and participates in TIA and IEC standards groups on optical fiber. He can be reached at (770)798-2737 or e-mail: firstname.lastname@example.org.